Vous êtes sur la page 1sur 22

OpenSeesPL 3D Pile-Ground Interaction

Jinchi Lu
University of California, San Diego
jinlu@ucsd.edu

Ahmed Elgamal

Zhaohui Yang

Thomas Shantz

September 3, 2010

Outline
Soil constitutive models
Solid-fluid fully-coupled elements
OpenSeesPL graphical user interface
Examples of using OpenSeesPL
Ongoing research projects

Pressure-Independent Material
Good for undrained clay
Nonlinear hysteretic behavior
and permanent deformation

PressureIndependMultiYield
nDmaterial PressureIndependMultiYield $tag $nd $rho $refShearModul
$refBulkModul $cohesi $peakShearStra <$frictionAng=0. $refPress=100.
$pressDependCoe=0. $noYieldSurf=20 <$r1 $Gs1 > >

Suggested parameter values:


Soft Clay

Medium Clay

Stiff Clay

1.3 ton/m3 or
1.217x10-4
(lbf)(s2)/in4

1.5 ton/m3 or
1.404x10-4
(lbf)(s2)/in4

1.8 ton/m3 or
1.685x10-4
(lbf)(s2)/in4

refShearModul

1.3x104 kPa or
1.885x103 psi

6.0x104 kPa or
8.702x103 psi

1.5x105 kPa or
2.176x104 psi

refBulkModul

6.5x104 kPa or
9.427x103 psi

3.0x105 kPa or
4.351x104 psi

7.5x105 kPa or
1.088x105 psi

18 kPa or
2.611 psi

37 kPa or
5.366 psi

75 kPa or
10.878 psi

0.1

0.1

0.1

rho

cohesi
peakShearStra

Pressure-Dependent Material
Multi-yield surface plasticity model (based on Prevost 1985)

Conical yield surfaces for granular soils (Prevost 1985; Elgamal et al.
2003; Yang and Elgamal 2008)

Pressure-Dependent Material (cont)


Incorporating dilatancy and cyclic mobility effects

Shear stress-strain and effective stress path under undrained shear loading
condition (Parra 1996, Yang 2000, Yang and Elgamal 2002)

PressureDependMultiYield
nDMaterial PressureDependMultiYield $tag $nd $rho $refShearModul
$refBulkModul $frictionAng $peakShearStra $refPress
$pressDependCoe $PTAng $contrac $dilat1 $dilat2 $liquefac1
$liquefac2 $liquefac3 <$noYieldSurf=20 <$r1 $Gs1 > $e=0.6 $cs1=0.9
$cs2=0.02 $cs3=0.7 $pa=101 <$c=0.3>>

Suggested parameter values:


Loose Sand
(15%-35%)

Medium Sand
(35%-65%)

Medium-dense
Sand (65%-85%)

Dense Sand
(85%-100%)

1.7 ton/m3 or
1.59x10-4
(lbf)(s2)/in4

1.9 ton/m3 or
1.778x10-4
(lbf)(s2)/in4

2.0 ton/m3 or
1.872x10-4
(lbf)(s2)/in4

2.1 ton/m3 or
1.965x10-4
(lbf)(s2)/in4

refShearModul (at
pr=80 kPa or 11.6
psi)

5.5x104 kPa or
7.977x103 psi

7.5x104 kPa or
1.088x104 psi

1.0x105 kPa or
1.45x104 psi

1.3x105 kPa or
1.885x104 psi

refBulkModu (at
pr=80 kPa)

1.5x105 kPa or
2.176x104 psi

2.0x105 kPa or
2.9x104 psi

3.0x105 kPa or
4.351x104 psi

3.9x105 kPa or
5.656x104 psi

frictionAng

29

33

37

40

peakShearStra (at
pr=80 kPa)

0.1

0.1

0.1

0.1

rho

Solid-Fluid Fully-Coupled Elements


(u-p formulation)
quadUP
9_4_QuadUP
brickUP
20_8_BrickUP

2D Solid-Fluid Coupled Element


Solid nodes: describe
the solid translational
degrees of freedom
Fluid nodes: describe
the fluid pressure

quadUP
element quadUP $eleTag $iNode $jNode $kNode $lNode $thick $matTag
$bulk $fmass $hPerm $vPerm <$b1=0 $b2=0 $t=0>

Recorder examples:
Pore water pressure:
recorder Node <-file $fileName> <-time> <-node ($nod1 $nod2 )> -dof 3 vel

Stress:
recorder Element <-file $fileName> <-time> <-ele ($ele1 $ele2 )> material 1 stress

2D Solid-Fluid Coupled Element


Solid nodes: describe
the solid translational
degrees of freedom
Fluid nodes: describe
the fluid pressure

9_4_quadUP
element 9_4_QuadUP $eleTag $Node1 $Node2 $Node3 $Node4 $Node5
$Node6 $Node7 $Node8 $Node9 $thick $matTag $bulk $fmass $hPerm
$vPerm <$b1=0 $b2=0>

Recommended to ensure numerical stability in case of


liquefaction analyses (Babuska-Brezzi condition)

3D Solid-Fluid Coupled Brick Element


Solid nodes: describe
the solid translational
degrees of freedom
Fluid nodes: describe
the fluid pressure

brickUP
element brickUP $eleTag $Node1 $Node2 $Node3 $Node4 $Node5
$Node6 $Node7 $Node8 $matTag $bulk $fmass $PermX $PermY
$PermZ <$bX=0 $bY=0 $bZ=0>

3D Solid-Fluid Coupled Brick Element


Solid nodes: describe
the solid translational
degrees of freedom
Fluid nodes: describe
the fluid pressure

20_8_BrickUP
element 20_8_BrickUP $eleTag $Node1 $Node20 $matTag $bulk
$fmass $PermX $PermY $PermZ <$bX=0 $bY=0 $bZ=0>

20-node element is recommended to ensure numerical stability in


case of liquefaction analyses (Babuska-Brezzi condition)

Soil Model Calibration


The pressure-dependent material model has been extensively
calibrated for clean Nevada Sand at Dr 40% (Parra 1996; Yang
2000)
Calibration was based on results of:
Monotonic and cyclic laboratory tests (Arulmoli et al. 1992)
Centrifuge experiments (VELACS Models 1 & 2, Dobry et al.
1995; Taboada 1995)

VELACS Models

OpenSeesPL Graphical User Interface

Soil Island (Soil Strata Only)

List of Soil Types

Pile in Soil Layers

Pile Properties

Pushover Analysis

Mesh Generation

10

Deformed Mesh

Pile Response Profiles & Time Histories

11

Soil Response Time Histories

Example 1: Pile in Liquefied Soil

0.2

b) 260-element mesh

10

0.1

10

0.1

Displacement (m)

0.2
1

Pile head longitudinal displacement


140element mesh
260element mesh
500element mesh

0
0

2m depth

0.5

c) 500-element mesh

20

Base excitation

Excess pore pressure (kPa)

Acceleration (g)

a) 140-element mesh

6
Time (sec)

10

12

20
60
40
20
6m depth

0
20
0

6
Time (sec)

10

12

Excess pore pressure time histories


in the free-field (500-element mesh)

12

Example 2: Stone Column Ground Modification

Schematic view
Of stone column
or pile layout

25

26

13

Example 3: Large Pile Group


Dumbarton Bridge

16 m 14 m

27 m

Pier 23

4 x 8 4.5 ft Diameter Hollow RC Piles


2D (long) x 2.15D (trans) pile spacing

8 x 4 Pile Group in Half Mesh

14

C (kPa)
34
58
74

G (MPa)
19
35
60

AppliedLoads

2.Pushoveranalysis(1g)

2.5
Pile group shear load (kN)

1.BridgeOwnweight

x 10

1.5

0.5

0.02

0.04
0.06
0.08
Pile cap displacement (m)

0.1

0.12

15

Front Corner Pile


Moment =1.35 of inner pile

Case of force only

Axialforce

Downdrag
Axialforce

16

Summary for Pile Group Study


For the employed cohesive soil scenario:
- Corner piles are most vulnerable
- Edge piles carry much more axial load than inner piles
- Axial response and skin friction (t-z springs) may play a critical
role
- Pile tensile forces may be quite significant and affect RC
response (connection to pile-cap, reduction in confinement)

Parallel Computing on Supercomputer


Machine: Teragrid IA-64

17

Ongoing Project: OpenSeesPBEE


Performance-Based Earthquake Engineering
(PBEE) Analysis of Bridge Systems - with Prof.
Mackie (Univ. of Central Florida)

Bridge and Ground Model

18

Input Motions

Concurrent Execution for Multiple Records

19

Performance Group (PG) Quantities

Report Generator

20

Unit Costs

PBEE Outcome: Repair Cost

21

Total Repair Cost Ratio

30% of original cost

Ongoing Project: Deep Soil Mixing


Joint project with UC Davis (Prof. Boulanger)
and Oregon State Univ. (Prof. Ashford)

22